Hi Luca, I've tried with different antennas which improved the level of the signal to -45 dBm sometime but I'm still facing the RF problem. Let me describe better the scenario: a) AP close (10cm not 10m) to the chip with antenna The signal is very good, I can connect to the AP on a unprotected network or WPA-AES protected one. Then I can ping the gateway steadily without any problem b) AP far (5m) from the chip with antenna In this case if I run: ./iw dev wlan0 scan It can happen that for 15 or 20 times I can't retrieve any network always executing the same command. After that, it can happen that for 15 or 20 times I can see one or both the networks available from the AP where the signal could be also -45 dBm. Then it can happen again that for 15 or 20 times I can't retrieve any network when scanning again. If I try to connect to the network (unprotected) when they are available I can't and I obtain: root@(none):~# ./iw wlan0 connect -w Inpeco-test wlan0 (phy #0): failed to connect, status: 1: Unspecified failure Where would you suggest us to look in order to overcome this problem we are fighting against ? Thanks in advance. Alberto -----Original Message----- From: Luciano Coelho [mailto:coelho@xxxxxx] Sent: lunedì 17 dicembre 2012 12:18 To: Alberto Garau Cc: linux-wireless@xxxxxxxxxxxxxxx; Ido Yariv; Arik Nemtsov; Adriano Melis; Luca Fresi Subject: Re: wl12xx over spi with no scan results On Mon, 2012-12-17 at 09:38 +0000, Alberto Garau wrote: > Hi Luca, Hi Alberto, > we have been able to successfully run the driver but: > > We have the device running in SPI and we are observing a strange scenario. When the AP is placed 10 meters far from the device (-57db) it's very difficult scanning the ESSIDs and impossible to connect to them, the same behaviour doesn't happen when I move the AP close (10 cm) to the device. In that case I can scan the ESSIDs and I can connect to them easily. This definitely sounds like an RF problem. Are you sure everything is correct with your antenna? 10 meters reach sounds like what happens if there's no antenna connected. ;) If you used the correct INI file, even without proper tuning of the settings, the RF shouldn't be that bad. In general, if you use the appropriate INI settings for your FEM, you should get relatively good connectivity. > The calibration we have performed is the First Calibration listed on that web site and we have more questions: > > 1) is there any document explaining in depth all the parameters of the > ini file TQS_S_2.6.ini You should get this from your official TI contact. Do you have any support channel? > 2) we used "./calibrator plt calibrate", what plt means ? PLT means Production Line Tests (calibration is usually done at the production line, so it's included with the rest of the PLT tests). > 3) if we want to use the expert mode, we would have to provide <voltage>,<power>,<subband> and then tx_bip parameters, where we can found indication/documentation about that ? This is used if you want to fine tune the RF path. To use it you need special equipment to test and measure the output and input. > 4) is there any auto calibration ? Yes, the calibration tool has the auto-calibration feature. > 5) Could you please define the meaning of the "TxBip procedure" and when should be used ? This is the procedure used to calibrate the device and generate some parameters that go into the NVS file. > 6) Could you please define the meaning of the "Tx Cont Procedure" and when should be used ? This is used in R&D and in the production line. It is used to transmit packets continuously using the parameters specified. > 7) Could you please define the meaning of the "Rx Stat procedure" and when should be used ? This command is used to get statistics about the received packets. It can be used to check whether the device is receiving packets as it should. Generally used in R&D and in the production line. I hope this helps. -- Luca. ��.n��������+%������w��{.n�����{���zW����ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
